Home / Research

How much does the debt snowball cost? A 10,000-portfolio simulation

A research note from WhenDebtFree, version 1.0.

Reviewed November 21, 2026

Summary. Across 10,000 simulated households, paying debts smallest-balance-first (snowball) cost a median $196 more in interest than highest-rate-first (avalanche). The middle half of households paid between $39 and $527 more; one in ten paid over $999 more. The gap is driven mainly by the spread between a household’s highest and lowest interest rates: a median $37 when rates are within 5 points of each other, and $670 when they are 20 or more points apart.

Key findings

  1. The typical cost of snowball is modest. The median extra interest was $196, or 5.4% of the interest paid under avalanche. 37% of households paid less than $100 more, and 10% paid more than $1,000 more.
  2. Snowball never came out ahead. In none of the 10,000 portfolios was snowball cheaper; it cost the same or more in every case.
  3. Interest rate spread is the main driver. The median extra cost rises from $37 (1.4% of interest) at a spread under 5 points to $670 (19.3%) at 20 points or more, where 32% of households paid over $1,000 extra.
  4. More debts, bigger gap. With two debts the median difference was $4, because both methods often choose the same order. With six debts it was $466.
  5. The cost is mostly money, not time. 67% of households finished in the same month either way; the median difference was 0 months, and the 90th percentile 1 month.

Results

How much more snowball costs, share of 10,000 portfolios
0%10%20%30%18%Under $1019%$10–10019%$100–25018%$250–50016%$500–1k9%$1k–2.5k1%Over $2.5kExtra interest paid with snowball instead of avalanche
Median extra cost of snowball, by interest rate spread
$0$200$400$600$800$1,000$1,200$1,400$370-5 pts$1805-10 pts$33310-15 pts$48815-20 pts$67020+ ptsSpread between highest and lowest APR (percentage points)

Bars show the median; whiskers show the 25th to 75th percentile.

By interest rate spread

APR spreadPortfoliosMedianMiddle half90th pct.Share of interestOver $1,000
0-5 pts2,616$37$0–$109$2141.4%0%
5-10 pts2,722$180$64–$343$5854.4%2%
10-15 pts1,490$333$118–$614$9848.6%9%
15-20 pts1,623$488$174–$852$1,33714.2%19%
20+ pts1,549$670$304–$1,160$1,81919.3%32%

By number of debts

DebtsPortfoliosMedianMiddle half90th pct.Share of interestOver $1,000
22,003$4$0–$112$3830.2%3%
32,017$95$13–$297$6353.5%4%
41,999$230$75–$532$9686.5%9%
51,963$345$150–$700$1,1867.7%14%
62,018$466$216–$850$1,4298.2%19%

By monthly budget

Tighter budgets stretch the payoff out, so the dollar gap is larger, though the share of interest changes little.

Budget vs. minimumsPortfoliosMedianMiddle half90th pct.Share of interestOver $1,000
1.0–1.5×2,484$308$59–$769$1,3994.5%18%
1.5–2.0×2,492$231$55–$605$1,0915.8%12%
2.0–2.5×2,469$173$31–$442$8135.7%6%
2.5–3.0×2,555$137$30–$366$6775.7%4%

Method

We generated 10,000 synthetic debt portfolios and paid each one off twice, once with each method, under identical conditions. The random seed is fixed (20261121), so rerunning the published code reproduces every figure on this page exactly.

  • Number of debts: 2 to 6, equally likely.
  • Balances: log-normal with a median of $3,500 (σ = 0.9 on the log scale), limited to $500–$25,000.
  • Interest rates: 70% drawn from a normal distribution centered on 22% (standard deviation 2.5 points), 30% drawn uniformly from 0–29%, then limited to 0–29%. This weights rates toward the 18–26% range typical of card balances while including low and promotional rates.
  • Monthly budget: the sum of the starting minimum payments multiplied by a random factor between 1 and 3. The budget stays fixed until every debt is paid, so freed-up minimums roll into the next target.
  • Payoff rules: the same as every calculator on this site, set out on our methodology page: monthly interest at APR ÷ 12, minimums of that month’s interest plus 1% of the balance with a $25 floor, and remaining money sent to the target debt.
  • Order: avalanche pays the highest rate first (ties: smaller balance); snowball pays the smallest current balance first (ties: higher rate).

Limitations

  • The portfolios are synthetic. They follow realistic ranges but are not drawn from real borrower records.
  • Rates are fixed, and there are no new purchases, fees, promotional periods or missed payments.
  • The study measures cost only. Snowball’s case rests on motivation, paying off whole accounts sooner, which a simulation can’t capture. A household that sticks with snowball will usually do far better than one that abandons avalanche.

Data and code

Everything needed to check or extend this study is published:

  • simulate_snowball_vs_avalanche.pyThe full simulation in standard-library Python 3. Run it to reproduce both data files.
  • summary.csvAggregated results: every table on this page.
  • portfolios.csvOne row per simulated portfolio: debts, balances, budget, rate spread, interest and months under each method.

The data and code may be reused, including commercially, with attribution under the Creative Commons Attribution 4.0 license.

How to cite

WhenDebtFree (2026). How much does the debt snowball cost? A 10,000-portfolio simulation (version 1.0). https://whendebtfree.com/research/snowball-vs-avalanche-cost/

Questions or a possible error? See our corrections policy. Try both methods on your own debts with the free calculator, or read snowball vs. avalanche explained.